Structure-activity relationships for analogues of the phenazine-based dual topoisomerase I/II inhibitor XR11576

Article abstract of DOI:10.1016/S0960-894X(01)00770-3

As part of a programme to identify further analogues of the dual topo I/II inhibitor XR11576, we describe here the syntheses and SAR studies of various 'minimal' and 3,4-benzofused phenazine chromophores of the phenazine template of XR11576.

Full text of DOI:10.1016/S0960-894X(01)00770-3

Bioorganic & Medicinal Chemistry Letters 12 (2002) 415–418
Structure–Activity Relationships for Analogues of the
Phenazine-Based Dual Topoisomerase I/II Inhibitor XR11576
Shouming Wang,a,* Warren Miller,a John Milton,a Nigel Vicker,a Alistair Stewart,b
c
Peter Charlton,b Prakash Mistry,b DavidHardick a andWilliam A. Denny
aDepartment of Medicinal Chemistry, Xenova Ltd., 957 Buckingham Avenue, Slough, Berkshire SL1 4NL, UK
bDepartment of Pharmacology, Xenova Ltd., 957 Buckingham Avenue, Slough, Berkshire SL1 4NL, UK
cAuckland Cancer Society Research Centre, Faculty of Medical and Health Sciences, The University of Auckland,
Private Bag 92019, Auckland, New Zealand
Received16 August 2001; revised2 November 2001; accepted13 November 2001
Abstract—As part of a programme to identify further analogues of the dual topo I/II inhibitor XR11576, we describe here the
syntheses andSAR studies of various ‘minimal’ and3,4-benzofusedphenazine chromophores of the phenazine template of
XR11576. # 2002 Elsevier Science Ltd. All rights reserved.
Considerable efforts have been devoted to the develop-
ment of dual inhibitors of the enzymes topoisomerase I
andtopoisomerase II that play vital roles in DNA
replication andtranscription, maintaining the desired
topology of DNA by catalyzing the breakage and
rejoining of strands.1 Topo I catalyzes the passage of
DNA strands through a transient single strand break,
while topo II catalyzes the passage of DNA double
strands through a transient double strand break.2,3
During the process, a covalent linkage is formed
between topoisomerase andDNA, termedas ‘cleavable
complexes’.4 The majority of topoisomerase inhibitors
act as poisons by stabilizing the ‘cleavable complexes’
more, topo I expression shows no major changes during
the cell cycles, whereas topo II expression increases in
response to DNA replication, reaches a maximum at the
G2-M phase and decreases rapidly after that.11 The
expression of either enzyme appears to be sufficient to
support cell division and the development of resistance
to topo I inhibitors is often accompaniedby a con-
12ꢀ14
comitant rise in the level of topo II andvice versa.
Dual inhibitors of topo I andII may thus have advan-
tages andthere has been much interest in such com-
pounds in clinical evaluation, including intoplicine15 (1)
16
andXR 5000
new phenazine-baseddual inhibitor, XR11576
(2). More recently, we have reporteda
17
5
andforming ternary complexes andhinder a specific
(3), a
step in the catalytic cycle, that is the reclosure of DNA
breaks. In contrast, direct inhibition of topoisomerase
catalytic activity without DNA breakage is also
known.6 Topoisomerase II was initially recognizedas an
important therapeutic target with the development of a
number of topo II inhibitors, such as doxorubicin and
etoposide,7 while camptothecin analogues were shown
to inhibit topo I.8
development candidate which circumvents multi-drug
resistance (MDR) and atypical MDR due to the down-
regulation of topo II.18
Topo I andtopo II are expressedat different absolute
levels in different cell types; the expression of topo I is
9
markedin colon cancer cell lines, while that of topo II
10
is high in many breast andovarian cell lines. Further-
*Corresponding author. Tel.: +44-1753-706768; fax: +44-1753-
706607; e-mail: shouming_wang@xenova.co.uk
0960-894X/02/$ - see front matter # 2002 Elsevier Science Ltd. All rights reserved.
PII: S0960-894X(01)00770-3

416
S. Wang et al. / Bioorg. Med. Chem. Lett. 12 (2002) 415–418
Herein we report a series of synthetic andstructure–
activity studies around this new phenazine-based lead
compound.
The Rf value observedindicates that the active com-
pounds are not substrates for P-gp. The 40-substituted
analogue (12b) appears to offer some scope for improv-
ing the activity over the parent analogue (12a), as
19
observedby Denny
in 2-phenyl quinoline-8-carbox-
amide series; compound 12c is twofoldmore potent
than analogue 12a andthis can be attributedto the
presence of a larger chromophore. By analogy, these
data clearly suggest that the tricyclic carboxamides not
possessing a complete fusedchromophore can still act
as the ‘minimal’ DNA intercalating agents for cytotoxic
activity. Indeed, severe steric crowding, as in compound
13, wouldprevent chromophore planarity, leading to
complete loss of activity. Predictably, the 2,3-naphthyl
fusedanalogue ( 14) with a large planar chromophore
shows a 10-foldincrease in potency.
‘Minimal Chromophore’ 3-Aryl Quinoxaline-5-carbox-
amides (6)
Previous work has demonstrated that suitably sub-
stituted2-aryl quinoline-8-carboxamides ( 5) can act as
the ‘minimal chromophore’ for DNA intercalation.19
A
broadspectrum of anti-tumor activity (in vivo) has been
shown for these compounds due to their superior dis-
tributive properties. This promptedus to assess the
corresponding phenazine derived chromophore (6) as a
potential series with desired physicochemical properties
for gooddistribution.
It is clear that the levels of activity in the ‘minimal
chromophore’ series are significantly less than those in
the 8,9-benzofusedphenazine series (XR11576,
17
IC50=20 nM in both H69/P andH69/LX4). Without
further enlargement of the chromophore (as in the pen-
tacyclic 14), the potential to achieve the same level of
potency appears to be limited, and therefore, consider-
ing the potential metabolic liability of pentacyclic com-
pounds such as 14, we quickly movedon to another
series with a tetracyclic chromophore.
The preparation of 3-aryl quinoxaline-5-carboxamides
(6) is outlinedin Scheme 1.
Aryl glyoxals (8) were preparedby heating a-bromo-
acetoketones (7) and48% HBr in DMSO at 50 ꢁC for
24 h.20 The subsequent regioselective condensation with
2,3-diaminobenzoic acid (9)17 in ethanol gave the quin-
oxaline acids (10) as the major products. The regio-
selectivity observedin this reaction was confirmedby
NOESY experiment on 10 (R=H). The acids (10) were
then coupledwith diamine ( 11) in the presence of CDI
to afford the amide products (12) in goodyields. Initi-
ally, the cytotoxic activity of these compounds was
measuredusing the H69 parental (H69/P) human small
cell lung carcinoma cell line andthe multi-drug resistant
human small cell lung carcinoma cell line (H69/LX4),
which over-expresses P-glycoprotein (P-gp).17 These
H69 cell lines represent typical examples of solidtumor
cell lines.
3,4-Benzofused Phenazine Chromophore: Benzo[a]-
phenazine-6-carboxamides (15)
Our recent development candidate, XR11576, was
developed through the chromophore addition by fusing
the benzene ring to the 8,9 positions of the DNA inter-
calating phenazine template (4).21 The angular 8,9-ben-
zofusedphenazine tetracyclic chromophore overlays
well with that of intoplicine (1). Meanwhile, we were
also interestedin exploring further another angular tet-
racyclic chromophore (15), by fusing the benzene ring to
the 3,4 positions of the phenazine template (4), as the
naked3,4-benzofusedphenazine-6-carboxamied was
shown to be a potent cytotoxin against P388 andLewis
lung carcinoma cell lines.21
The activity of selectedexamples of the quinoxaline
analogues is shown in Table 1. These data illustrate that
inhibitory activity is observedin both cell lines with
nearly equal potency for all the compounds examined.
Table 1. Cytotoxic activity of quinoxaline carboxamides (12)
CompdR
R2
H69/P
IC50 mM
H69/LX4
IC50 mM
Rf
LX4/Pa
1
12a
12b
12c
13b
Ph
H
H
H
Ph
3.93
2.40
1.82
>100
3.59
3.38
1.67
>100
0.9
1.41
0.92
4-(NO2)Ph
2-Naphthyl
Ph
14b
0.35
0.33
0.93
aRf LX4/P: the ratio of the IC50 value on the H69/LX4 cell line over
the IC50 value on the H69 parental cell line.
bCompounds 13 and 14 were preparedvia the condensation of the
respective diketone with 2,3-diaminobenzoic acid (9).
Scheme 1. Reagents andconditions: (a) DMSO, 48% HBr, 50 ꢁC,
46%; (b) EtOH, reflux, 93%; (c) CDI, DMF, 11, rt, 56%.

S. Wang et al. / Bioorg. Med. Chem. Lett. 12 (2002) 415–418
417
The general synthesis of (substituted) benzo[a]-phena-
zine-6-carboxamides is outlined in Schemes 2 and 3.22
sequent amide formation gave products (21). The cor-
responding 3-hydroxy analogue (20e, R=3-OH) was
also reactedwith (thio)isocyanates to give products ( 22).
For 8-methyl substitutedbenzophenazine-6-carbox-
amides (26), the chemistry described in Scheme 3 was
developed. Hence, treatment of 2-hydroxy-naphthoic-3-
acids (17) with sodium nitrite and 2 N HCl gave the
hydroxyimino-oxo acid (23)23 in goodyields. The sub-
sequent condensation with 2,3-diaminotoluene (24)
afforded the 8-methyl substituted phenazine acids (25)
regioselectively, due to a combination of relative func-
tional reactivity of 23 andsteric effects of the nucleo-
philic amino groups in 24. The formation of the amides
was achievedusing the standardprotocols. A NOESY
experiment was carriedout on 26, (R=H) andits
regiochemistry was confirmed.
When a mixture of benzofuroxan (16) and2-hydroxy-
naphthoic-3-acidor its substitutedanalogues ( 17) was
treatedwith 10% NaOH, the benzophenazine id- N-
oxides (18) were formedin goodyields. The reduction of
the dioxides (18) was accomplishedusing sodium thio-
sulfate in 0.5 N NaOH, leading to the key intermediates
(19). Subsequent coupling with the diamine (11) using
the standard protocols then afforded a number of (sub-
stituted) amides (20). Further alkylations of the 3-
hydroxy intermediates (19) with alkyl halides and sub-
Table 2 shows some of the cytotoxic activity data on
selectedexamples.
As the data in Table 2 illustrate, these compounds are
much more cytotoxic than those in Table 1, as expectedfor
agents likely to have tighter DNA binding (although the
mechanism of this cytotoxicity has not been determined).
The parent analogue (20a) shows almost equal activity
against H69/P andH69/LX4 cell lines with potency
around500 nM; therefore it is not pumpedby P-gp.
Nitro group substitution at the 1-position does not alter
the activity considerably (20b), whereas the electron
donating amino analogue (20c) gives slight better activ-
ity. The 3-methoxy substitutedanalogue ( 20d) is as
active as 20c in both cell lines, but the 3-hydroxy analo-
gue (20e) shows more than two-foldenhancement in
activity over 20d, although a slight increase in resistance
is observed( Rf=2.12). This suggests that it may be
Table 2. Cytotoxic activity of benzo[a]-phenazine-6-carboxamide
analogues
Scheme 2. Reagents andconditions: (a) 10% NaOH, rt, 16 h, then
HCl, 63–90%; (b) Na2S2O3, 0.5 N NaOH, 100 ꢁC, 15 min, then AcOH,
78%; (c) CDI, 11; (d) NaOH, H2O, EtOH, R0CH2Br; (e)
R00N¼C¼O(S),CH3CN/DCM.
CompdR
R2
H69/P
H69/LX4
Rf
1
IC50 mM IC50 mM LX4/P
20a
20b
20c
20d
20e
H
H
H
H
H
H
0.406
0.458
0.259
0.282
0.094
0.511
0.584
0.345
0.358
0.198
1.26
1.30
1.33
1.30
2.12
1-NO2
1-NH2
3-OMe
3-OH
21a
H
0.435
0.613
1.40
21b
21c
21d
3-O(CH2)2OMe
3-O(CH2)3OH
3-OBn
H
H
H
0.445
0.523
1.223
0.545
2.026
1.581
1.23
3.87
1.29
21e
H
0.506
0.592
1.17
22a
22b
26
3-OCONHPh
3-OC(S)NHEt
H
H
H
8-CH3
0.104
0.519
0.417
0.252
0.583
0.320
2.42
1.12
0.80
Scheme 3. Reagents andconditions: (a) NaOH, H 2O, then NaNO2,
2 N HCl, 10 ꢁC, 69%; (b) AcOH, concdHCl, 100 ꢁC, 66%; (c) CDI,
DMF, 11.

418
S. Wang et al. / Bioorg. Med. Chem. Lett. 12 (2002) 415–418
pumpedby P-gp to a small extent. It has been proposed
21
References and Notes
by Vicker et al.17 andRewcastle that biological activ-
ity in this class of compounds depends critically not
only on intercalation of the chromophore but also on
positioning of the carboxamide side chain. However,
investigation for finding further potential binding inter-
actions in 3,4-benzofusedregions is relatively unknown.
Therefore, a strategy was devised to derivatize the 3-
hydroxy group (20e) with a number of functional
groups with the aim of discovering any potential bind-
ing sites. These groups included the propargyl group
(21a), a proton acceptor methoxy ethyl group (21b) and
a proton donor group (21c), which all give decreased
activity relative to 20e. Notably, compound 21c that
bears a hydroxy group in the side chain, like 20e shows
some degree of resistance (Rf=3.87). Introduction of
the solublilizing morpholine group fails to show any
improvement in activity (21e), while analogue 21d, that
bears a lipophilic benzyl group, results in a 10-foldloss
of activity comparedwith 20e. Different linker groups
such as carbamate groups were also examined, with
compound 22a showing the same level of potency as
20e. In contrast, the ethyl thio carbamate analogue
(22b) showeda loss of potency.
1. Wang, J. C. Annu. Rev. Biochem. 1985, 54, 665.
2. Ferro, A. M.; Olivera, B. M. J. Biol. Chem. 1984, 259, 547.
3. Muller, M. T.; Spitzher, J. R.; Didonato, J. A.; Metha,
V. B.; Tsutsui, K. Biochemistry 1988, 27, 8369.
4. Nelson, E. M.; Tewey, K. M.; Liu, L. F. Proc. Natl. Acad.
Sci. U.S.A. 1984, 81, 1361.
5. (a) Robinson, M. J.; Osheroff, N. Biochemistry 1991,
30, 1807. (b) Burden, D. A.; Kingma, P. S.; Froelich-Ammon,
S. J..; Bjornski, M. A.; Patchan, M. W.; Thompson, R. B.;
Osheroff, N. J. J. Biol. Chem. 1996, 271, 29238.
6. Capranico, G.; Binaschi, M.; Evelina Borgnetto, M.;
Zunino, F.; Palumbo, M. Tips 1997, 18, 323.
7. Denny, W. A. Exp. Opin. Invest. Drugs 1997, 6, 1845.
8. Dancey, J.; Eisenhauer, E. A. Br. J. Cancer 1996, 74, 327.
9. Husain, I.; Mohler, J. L.; Seigler, H. F.; Besterman, J. M.
Cancer Res. 1994, 54, 539.
10. Holden, J. A.; Rolfson, D. H.; Wittwer, C. T. Biochem-
istry 1990, 29, 2127.
11. Heck, M. M. S.; Hittelman, W. N.; Earnshaw, W. C.
Proc. Natl. Acad. Sci. U.S.A. 1998, 85, 1086.
12. Tan, K. B.; Mattern, M. R.; Eng, W.-K.; McCabe, F. L.;
Johnson, R. K. J. Natl. Cancer Inst. 1989, 81, 1732.
13. Sugiomoto, Y.; Tsukahara, S.; Oh-Hara, T.; Isoe, T.; Tsu-
Ruo, T. Cancer Res. 1990, 50, 6925.
14. Whiteacre, C. M.; Zborowska, E.; Gordon, N. H.;
Mackay, W.; Berger, N. A. Cancer Res. 1425, 997, 57.
15. Poddevin, B.; Riou, J. F.; Lavelle, F.; Pommier, Y. Mol.
Pharmacol. 1993, 44, 767.
16. Finlay, G. J.; Riou, J. F.; Baguley, B. C. Eur. J. Cancer
1996, 32A, 708.
17. Vicker, N.; Burgess, L.; Chuckowree, I.S.; Dodd, R.;
Folkes, A.; Hardick, D. J.; Hancox, T. C.; Miller, W.; Milton,
J.; Sohal, S.; Wang, S.; Wren, S. P.; Charlton, P.; Dangerfield,
W.; Liddle, C.; Mistry, P.; Stewart, A.; Denny, W. A. J. Med.
Chem. In press.
Rewcastle has shown that introduction of a methyl
group peri to the phenazine N gives enhancement of
activity in their assay (P388 leukemia andLewis lung
carcinoma cell lines).21 However, in the 3,4-benzofused
series, no such effect is observedin our assays ( 26).
Although these data show good cytotoxic activity for
analogues with the 3,4-benzofusedphenazine chromo-
17
phore, it is clear that the 8,9-benzofusedseries
dis-
playedsuperior potency andless toxicity. This resulted
in our main biochemical focus being on the latter series,
culminating in the selection of the development candi-
date, XR11576.17
18. (a) Beck, W. T.; Cirtain, M. C.; Danks, M. K.; Felsted, R. L.;
Safa, A. R.; Wolverton, J. S.; Suttle, D. P.; Trent, J. M. Cancer
Res. 1987, 47, 5455. (b) Isaacs, R. J.; Davies, S. L.; Wells, N. J.;
Harris, A. L. Anti-cancer Drugs 1995, 6, 195.
19. Atwell, G. J.; Baguley, B. C.; Denny, W. A. J. Med. Chem.
1989, 32, 396.
20. Saggiomo, A. J.; Kano, S.; Kikuchi, T.; Okubo, K.;
Shinbo, M. J. Med. Chem. 1972, 15, 989.
Acknowledgements
21. Rewcastle, G. W.; Denny, W. A.; Baguley, B. C. J. Med.
Chem. 1987, 30, 843.
22. Ley, K.; Seng, F. Synthesis 1975, 415.
The authors wouldlike to thank Dr. MohammedLatif
for analytical support andDrs. Tim Hancox, Stephen
Wren and Mr. Adrian Folkes for proof reading.
23. Ullmann, H. Chem. Ber. 1909, 42, 4266.